We want to split the data into deciles on each data column. First, we’ll filter by the top decile for retailer concentration and retailer count. From there, we will count decile number as an integer to add across a tract. That will make a score for how extreme a tract’s case is for all demographic data points. Taking the top ten scoring tracts for every state will get us started narrowing down the worst hit tracts for reporters to focus on.
retailer_demo_map